Refresh ASI-Bench dataset card

README.md

@@ -1,53 +1,52 @@
 ---
 pretty_name: ASI-Bench Generated Instances (Seed 31415)
 language:
-  - en
+- en
 license: apache-2.0
 size_categories:
-  - n<1K
+- n<1K
 tags:
-  - ai-for-science
-  - benchmark
-  - agents
-  - scientific-reasoning
+- benchmark
+- agents
+- ai-for-science
+- scientific-reasoning
 ---
 
-# ASI-Bench Generated Instances
+<div align="center">
 
-For usage instructions, task definitions, evaluation details, and all other
-information, see the [ASI-Bench repository](https://github.com/apexin-ai/ASI-Bench).
+# ASI-Bench Generated Instances — Seed 31415
 
-## Dataset structure
+[Website](https://asibench.apexin.ai/) · [GitHub](https://github.com/apexin-ai/ASI-Bench) · [Leaderboard](https://asibench.apexin.ai/leaderboard)
+
+</div>
+
+## Dataset Summary
+
+[ASI-Bench](https://github.com/apexin-ai/ASI-Bench) evaluates general intelligence, innovation, and autonomous execution through 60 project-level scientific research tasks spanning 11 domains. Each task provides four matched prompt conditions: B1 gives full methodological guidance, B2 identifies the method without the full procedure, B3 provides only the research objective and data, and B4 adds task-irrelevant information to B3.
+
+This repository contains the fixed seed `31415` generated instances, including prompts, task inputs, reference artifacts, and available instance-generation metadata.
+
+> For complete benchmark information and instructions for downloading, running, and submitting evaluations, see the [ASI-Bench GitHub repository](https://github.com/apexin-ai/ASI-Bench).
+
+## Dataset Structure
 
 ```text
-.
-├── README.md
-├── LICENSE
-├── metadata.json
-├── .gitattributes
-└── tasks/
-    └── <instance-id>/
-        ├── prompt_b1.md
-        ├── prompt_b2.md
-        ├── prompt_b3.md
-        ├── prompt_b4.md
-        ├── data/
-        ├── reference/
-        ├── instance_meta.json
-        └── framework_task_info.json
+tasks/
+└── <task_id>__seed31415/
+    ├── prompt_b1.md
+    ├── prompt_b2.md
+    ├── prompt_b3.md
+    ├── prompt_b4.md
+    ├── data/
+    │   └── ...
+    ├── reference/
+    │   └── ...
+    ├── instance_meta.json          # where available
+    └── framework_task_info.json    # where available
 ```
 
-Each directory under `tasks/` is one generated benchmark instance.
-
-- `prompt_b1.md` through `prompt_b4.md` contain the four prompt variants.
-- `data/` contains the task input files.
-- `reference/` contains the reference answer and validation artifacts.
-- `instance_meta.json` records generation metadata.
-- `framework_task_info.json` records framework and re-evaluation metadata.
-- `metadata.json` is the dataset-level manifest.
+Each directory under `tasks/` is one generated benchmark instance. `data/` contains task inputs, `reference/` contains reference artifacts, and the JSON files record available generation and framework metadata.
 
 ## License
 
-Except where otherwise noted, this dataset repository is licensed under the
-[Apache License 2.0](LICENSE). Third-party components remain subject to their
-respective licenses and terms.
+Except where otherwise noted, this dataset repository is licensed under the [Apache License 2.0](./LICENSE). Third-party components remain subject to their respective licenses and terms.